Do alumni relationships between executive directors and auditors matter for financial reporting quality?

ABSTRACT

This paper examines the impact of executive director-auditor Guanxi on Chinese listed firms’ financial reporting quality from 2009 to 2019. We identify Guanxi-connected auditors as those who attend the same university as the client firm directors. Using a fixed-effect model, we provide consistent evidence that alumni ties between auditors and executive directors are positively related to corporate earnings management, indicating that auditors are more willing to trust clients and less likely to challenge their financial reporting decisions. Subsequently, client firms have more incentives to manipulate earnings, and financial reporting quality is reduced. Our study also finds that the positive impact of director-auditor ties on earnings management is greater among non-state-owned enterprises, firms that do not recruit big-four auditors, and those with fewer industry specialist auditors. Our study sheds new light on the impact of Guanxi-connected auditors on financial reporting quality, enhancing our understanding on the role of social ties.

Highlights

• Alumni ties between auditors and executive directors are positively related to earnings management.

• Such a positive relationship is greater among non-state-owned enterprises.

• Such a positive relationship is more pronounced for firms that do not recruit big-four auditors.

• Such a positive relationship is more pronounced for firms with fewer industry specialist auditors.

KEYWORDS:

• Auditors
• earnings management
• executive directors
• financial reporting quality
• Guanxi
• social ties

Acknowledgement

We are grateful for the feedback provided by Associate Editor Zhongwei Huang and the anonymous referees during the review process

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We use the terms social ties, Guanxi ties, educational ties, social networks interchangeably throughout the paper.

2 On the one hand, when there is an existence of “alumni relationships”, directors are more likely to pressure accountants to adopt more aggressive accounting approaches to reach the desired numbers, as they are aware that the likelihood of receiving modified audit opinions can be significantly reduced. As a result, more favorable audit outcomes lead to a reduced likelihood of firms being investigated or punished (Liu et al., 2011). On the other hand, accountants, who succumb to such pressure, tend to adopt more aggressive practices to recognize revenues or expenses and are forced to do “whatever was necessary” to meet the earnings targets and managers’ expectations (Albrecht et al., 2015).

3 There are typically two signing auditors: one senior signing auditor takes charge of the decisions in the auditing process, and the junior conducts the fieldwork.

4 The estimates of $\widehat{{\alpha }_0}$, $\widehat{{\alpha }_1}$, $\widehat{{\alpha }_2}$ are obtained from the original Jones Model.

5 We calculate industry specialization for the lead audit firm through dividing a client firm’s total assets by the sum of total assets of all firms within the same industry. The computation of AIS is proceeded as follows: $AI{S}_{ik}=\sum _{j=1}^J\sqrt{Asse{t}_{ikj}}/\sum _{i=1}^I\sum _{j=1}^J\sqrt{Asse{t}_{ikj}}$ where $\sum _{j=1}^J\sqrt{Asse{t}_{ikj}}$ is the sum of squares of a client firm i's total assets corresponding to the lead audit firm j in an industry k.

6 We have also used a correlation matrix to examine the multicollinearity issue and the results are available upon request. The Pearson correlation coefficients among all variables are less than 0.47, indicating that the multicollinearity issue is not an issue in our study.

7 There are concerns regarding the use of a two-step model of estimating discretionary accruals. In a two-step model, we first estimate the discretionary accruals and then regress it on a set of second-stage variables without adding the first-step regressors. However, this two-step regression procedure may generate biased estimates regardless of the estimation form, leading to Type I and Type II errors (Chen et al., 2018). Following Chen et al. (2018), we include the first-step regressors in our second-step model and re-estimate the model to address such concerns. Appendix Table 2 presents the results of the revised model. It shows that our main findings remain unchanged.

8 We select 1-to-3 matching because it can yield higher precision than 1-to-1 matching (Rassen et al., 2012). Having said this, we have done from 1-to-1 to 1-to-4 matching, and our results remain consistent.

9 We collect financial misconduct data from the CSMAR database.

10 We identify that executive directors and auditors are in the similar age range if their age differences are within two years. In this case, they are more likely to become classmates.

11 The Chinese government launched “211 Project” in 1995. The universities included in “211 Project” list are the most reputable universities in China. To further check the robustness of our analysis, we also include the accounting and finance related universities whose rankings are comparable to the “211 Project” universities. The results (untabulated) remain unchanged.

12 For brevity, we do not report the first-stage results for sub-sample analysis in Table 9. The results are available upon request.